INFLUENCE OF PATIENT AGE AND COLORECTAL POLYP SIZE ON HISTOPATHOLOGY FINDINGS

Background Colorectal cancer is a major cause of morbidity and mortality and can arise through the adenoma-carcinoma sequence. Colonoscopy is considered the method of choice for population-wide cancer screening. Aim To assess the characteristics of endoscopically resected polyps in a consecutive series of patients who underwent colonoscopy at a university hospital and compare histopathology findings according to patient age and polyp size. Methods Retrospective, cross-sectional of 1950 colonoscopy reports from consecutively examined patients. The sample was restricted to reports that mentioned colorectal polyps. A chart review was carried out for collection of demographic data and histopathology results. Data were compared for polyps sized ≤0.5 cm and ≥0.6 cm and then for polyps sized ≤1.0 cm and ≥1.1 cm. Finally, all polyps resected from patients aged 49 years or younger were compared with those resected from patients aged 50 years or older. Results A total of 272 colorectal polyps were resected in 224 of the 1950 colonoscopies included in the sample (11.5%). Polyps >1 cm tended to be pedunculated (p=0.000) and were more likely to exhibit an adenomatous component (p=0.001), a villous component (p=0.000), and dysplasia (p=0.003). These findings held true when the size cutoff was set at 0.5 cm. Patients aged 50 years or older were more likely to have sessile polyps (p=0.023) and polyps located in the proximal colon (p=0.009). There were no significant differences between groups in histopathology or presence of dysplasia. Conclusion Polyp size is associated with presence of adenomas, a villous component, and dysplasia, whereas patient age is more frequently associated with sessile polyps in the proximal colon.


INTRODUCTION
C olorectal cancer (CRC) is a major cause of morbidity and mortality. It is the fourth most common malignant neoplasm and the third leading cause of cancer mortality in Brazil 1 . The incidence is higher between the ages of 50 and 70.
It is widely known that 60-90% of this cancer arise from adenomas 8  This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercia License, which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited. adenoma-carcinoma sequence. In the majority of cases, this transformation is relatively slow, taking up to 10-15 years 8 . This slow growth enables prevention of CRC by endoscopic resection of polyps.
In view of its prevalence, its long asymptomatic interval, and the presence of treatable precancerous lesions, CRC fulfills all criteria for routine population-wide screening. Colonoscopy is considered the method of choice for this purpose 20 . Randomized clinical trials and several cohort studies have shown that colonoscopic polypectomy reduces its incidence by 76-90%, as compared with a general population registry 22,29 .
Colorectal adenomas are the neoplasms most commonly detected during screening colonoscopy, as well as in diagnostic colonoscopy of symptomatic patients over the age of 50 28 . Adenomatous polyps may be classified as low-, moderate-, or high-risk lesions in terms of the risk of progression to cancer 29 . Lesions are considered advanced when they are ≥1 cm in size or exhibit a villous component or high-grade dysplasia 26 . Age is considered a risk factor for the presence of adenomas and dysplasia, the incidence of which increases once the sixth decade of life is reached 23 .
The objective of this study was to assess the characteristics of polyps resected endoscopically from a consecutive series of patients who underwent colonoscopy at a university hospital and compare histopathological findings by patient age and polyp size.

METHODS
This was a retrospective, cross-sectional, chart review study based on analysis of the reports of 1950 colonoscopies performed consecutively at the Coloproctology Service of Hospital Universitário de Brasília. The indications for colonoscopy were not taken into account. Reports were obtained from the hospital database. Only those that described evidence of polyps in the colon or rectum were considered for analysis. Patient charts were then reviewed to collect demographic data and the results of histopathological examination of resected specimens. Each polyp was analyzed individually, even when several were resected from the same patient.
Patients with inflammatory bowel disease, colorectal malignancy, or genetic syndromes associated with polyposis were excluded from the sample, as were incomplete colonoscopies, polyps with malignant transformation, and unresected polyps.
Initially, the histopathological features of resected polyps were compared according to polyp size, defined as a dichotomous variable (≤0.5 cm or ≥0.6 cm as estimated by the endoscopist). A second analysis then compared ≤1.0 cm and ≥1.1 cm polyps. Finally, all polyps, regardless of size, resected from patients aged 49 years or younger were compared with those resected from patients aged 50 years or older.
Polyps located proximal to the splenic flexure of the colon were considered proximal, whereas those located after the splenic flexure were distal.
Statistical analyses were performed in the SPSS 17.0 software environment. Fisher's exact test was used for betweengroup comparisons. The significance level was set at p<0.05.

RESULTS
A total of 272 colorectal polyps were resected in 224 of the 1950 colonoscopies included in the sample (11.5%). Most of these colonoscopies had been performed in women (55.1%), and 75.9% of patients were aged 50 years or older.
Polyps were solitary in 51% of cases. In terms of morphology, 79.8% were sessile and 20.2 % were pedunculated (stalked). The most frequent site was the left colon (43.4%), followed by the right colon (20.6%), the transverse colon (17.6%), and the rectum (17.6%). Polyps were scattered throughout the colon in 7% of cases. Most polyps were <1 cm in size according to the examining physician (88.6%). The polyp size estimated by the endoscopist matched that determined by the pathologist in 80.1% of cases.
Comparison of polyps by size, using 1 cm as a cutoff, showed that larger polyps tended to be pedunculated (p=0.000) and were more likely to exhibit an adenomatous component (p=0.001) and dysplasia (p=0.003). There were no betweengroup differences in the distribution of polyp sites (p=0.677, Table 1).
Histopathology findings according to polyp size are described in Table 2. Only 10.7% of adenomas ≤1 cm in size had a villous component, versus 56% of those larger than 1 cm (p=0.000).
Similar findings were observed when the size cutoff was set at 0.5 cm (Tables 3 and 4). Only 8.5% of adenomas ≤0.5 cm in size had a villous component, versus 57.6% of those 0.6 cm or larger (p=0.000).    Comparison of polyp histopathology according to patient age showed that subjects aged 50 or older were more likely to have sessile polyps (p=0.023) located in the proximal colon (p=0.009). There were no significant age-related differences in histopathology or presence of dysplasia (Table 5). Histopathological features according to patient age are described in Table 6. There were no statistically significant between-group differences in the presence of a villous component (p=0.511).

DISCUSSION
Colorectal polyps are common, being detected in up to 33% of colonoscopies 9 . Two-thirds of all colon polyps are adenomas, which, by definition, are dysplastic and have the potential for malignant transformation. Nearly all CRCs arise from adenomas, but only a small minority of adenomas will actually progress to cancer6.
The incidence of adenomatous polyps has been described as 21-28% in patients aged 50-59 years, 41-45% in the 60-69 age group, and 53-58% in patients over the age of 70 14 . The prevalence of adenomatous polyps on autopsy has been reported as 20-30%, and the incidence of these lesions appears to increase with age 19 . According to current ASGE/ACG recommendations, adenomas will be detected during first-ever colonoscopy in over 25% of asymptomatic men and 15% of asymptomatic women over the age of 50 20 .
The lower incidence of polyps in this study might be explained by the fact that the indication for colonoscopy was not taken into account, and that some of the colonoscopies included were performed under suboptimal bowel preparation conditions. The polyp detection rate depends on a host of variables, including the demographics of the screened population (age, sex, family history of CRS), the quality of bowel prep, endoscopist technique and expertise, and endoscope withdrawal time 9 .
Just over half of all polyps in this series (51%) were solitary. According to Lowenfels et al. 12 , approximately twothird of patients have solitary polyps, and the frequency of larger polyps increases with advancing age.
In this study, 91.9% of polyps <0.5 cm in size were sessile. Conversely, those larger than 1 cm were mostly pedunculated (67.7%). It is well known that polyps<5 mm, also known as minute polyps, are rarely stalked 6 .
Histopathological examination is accepted as the gold standard for definition of polyp size and has been recommended for clinical practice and research purposes alike 5 . In the present study, the polyp size estimated by the endoscopist at the time of resection matched the size later determined by the pathologist in 80.1% of cases. According to Schoen et al. 24 , polyp size is estimated inaccurately by the endoscopist in 20% of cases, with a trend toward overestimation. Conversely, other authors have concluded that endoscopists tend to underestimate lesion size 15 . In this study, polyp size was defined as that estimated by the endoscopist and noted in the colonoscopy report, so that histopathology findings could be interpreted from the point of view of the examining physician, who will be in charge of patient care and follow-up.
The histological features and size of adenomas are the most important determinants of malignant potential 6 . Adenomas may be classified as tubular, villous, or tubulovillous, according to their glandular architecture. Over 80% of colonic adenomas are tubular 16 .
Most polyps resected from the patients in this sample were ≤1 cm in size, left-sided, and had tubular adenoma as the predominant histopathological type, which corroborates previous findings 26 . However, in patients over the age of 50, polyps were most commonly located in the proximal colon. Prior studies have reported age as a major risk factor for proximal lesions 11 . Other authors, however, have found no age-related differences in polyp distribution 17 .
There was a higher incidence of adenoma and dysplasia in patients over the age of 50, but the difference did not reach statistical significance. Other studies have reported a higher incidence of adenomas in general and advanced adenomas in particular after the fifth decade of life 23,18 . There were no significant between-group differences in presence of villous component. Villous polyps may become malignant in 29-70% of cases 13 . The presence of a villous component in endoscopically resected adenomas is a predictor of advanced lesions on followup colonoscopy 28 .
Winawer et al. 29 , in an analysis restricted to polyps ≥1 cm in diameter, found that 86% of adenomas exhibited slight atypia, 8% were moderately atypical, and 6% showed marked atypia, also known as carcinoma in situ. In the present study, 54.7% of polyps ≥1 cm were slightly dysplastic, 22.6% were moderately dysplastic, and 3.3% exhibited high-grade dysplasia.
One important finding of this study was the absence of any significant difference in histopathology features when the size cutoff for polyps was set at 0.5 cm or 1.0 cm. In both cases, increasing polyp size was associated with increased odds of adenoma, villous component, and dysplasia. Therefore, one may conclude that small (6-9 mm) polyps should not be neglected.
Few studies have assessed the rate of advanced histology on the basis of polyp size 10 . One such study concluded that removal of a greater number of polyps (including smaller polyps) with a lower rate of adenoma resection is preferable to removal of fewer polyps for a higher rate of adenoma resection 3 .
Kim et al. 7 reported advanced histology in only 3% of polyps 6-9 mm in diameter. Other studies 30 found evidence of a villous component in 4-15% and high-grade dysplasia in 4.3-5.8% of polyps in this size range. Lieberman et al. 10 found a high proportion of advanced histology (prevalence up to 30.6%) in patients with polyps larger than 1 cm, whereas those with small (6-9 mm) polyps were at intermediate risk (6.6%), including of high-grade dysplasia (0.92%).
In another study 27 , which included patients aged 40-89 years, 18.7% of subjects had adenomas, 5% of which were advanced. The prevalence of advanced histology was 85% in polyps ≥1 cm, 27% in polyps 6-9 mm and 10% in polyps ≤5 mm in size. The authors concluded that failure to remove small polyps may place patients at risk of progression to advanced lesions and cancer.
Rex et al. 21 , in a retrospective study of 5079 patients, found advanced histology in 0.87% of minute (≤5 mm) polyps and 5.3% of small (6-9 mm) polyps. Chaput et al. 2 found advanced histology in 4.7% of minute and 35.2% of small polyps, mostly due to presence of a villous component. The authors noted that polyp size <1 cm was associated with a higher incidence of advanced adenoma.
In a retrospective study of patients undergoing first-ever colonoscopy, Shapiro et al. 25 found that 1.6% of polyps ≤5 mm exhibited high-great dysplasia or malignant transformation, and 4.1% contained a villous component. The rate of advanced histology for polyps 6-9 mm in size was over 15%. The authors concluded that expectant management of small polyps puts more than 5% of patients at risk of dysplasia progression.
In a systematic review by Hassan et al. 4 , advanced adenomas were identified in 5.6% of minute polyps, 7.9% of small polyps, and 87.5% of large (≥1 cm) polyps. The authors concluded that polypectomy of lesions larger than 6 mm identifies 95% of advanced adenomas. When resection is limited to polyps larger than 10 mm, only 88% of advanced lesions are identified.

CONCLUSION
Polyp size was associated with the presence of adenomatous and villous components and with dysplasia, whereas patient age was more frequently associated with sessile polyps located proximal to the splenic flexure.